What is Predictive Segmentation?

Predictive segmentation is the capability of identifying and creating new visitor segments to high-probability elements.

The task of segmenting and analyzing online traffic was very time consuming before the introduction of predictive analytics.

Identifying traffic segments manually can be done in two ways. One way is by a bottom-up approach that segments traffic using available data and the primary KPI, or alternatively one can use a top-down approach that identifies traffic based on the KPI then breaks it down per data point.

• Data-points often include:
• Geographic location
• Devise operational system data
• Browser software data
• Channel source of acquisition
• Are they are returning or a new visitor
• Demographics

To segment traffic, marketers must first sort through all of the traffic and identify the most valuable and underperforming segments. Once these groups have been identified, experience tailored to each group must be provided in order to maximize results.

Predictive segmentation serves to reduce manual labor by automatically identifying segments with an indication of high-potential traffic, but does not offer any specific details for what type of experience should be tailored to each identified segment. In other words, predictive segmentation identifies which segments are best to focus on without giving any guidance as to the specifics for the experience.